Stepping Up the Cybersecurity Game: Protecting Online Services from Malicious Activity

The rise in popularity of online services such as social networks,web-based emails, and blogs has made them a popular platform for attackers.Cybercriminals leverage such services to spread spam, malware, and stealpersonal information from their victims.In a typical cybercriminal operation, miscreants first infect their victims' machines with malicious software and have themjoin a botnet, which is a network of compromised computers. In the second step,the infected machines are often leveraged to connect to legitimate onlineservices and perform malicious activities.As a consequence, online services receive activity from both legitimate and malicious users. However, while legitimate users use these services for thepurposes they were designed for, malicious parties exploit them for theirillegal actions, which are often linked to an economic gain. In this thesis, I showthat the way in which malicious users and legitimate ones interact with Internetservices presents differences. I then develop mitigation techniques thatleverage such differences to detect and block malicious parties that misuseInternet services.As examples of this research approach, I first study the problem of spammingbotnets, which are misused to send hundreds of millions of spam emails tomailservers spread across the globe. I show that botmasters typically split alist of victim email addresses among their bots, and that it is possible toidentify bots belonging to the same botnet by enumerating the mailservers thatare contacted by IP addresses over time. I developed a system, calledBotMagnifier, which learns the set of mailservers contacted by the bots belongingto a certain botnet, and finds more bots belonging to that same botnet.I then study the problem of misused accounts on online social networks. I firstlook at the problem of fake accounts that are set up by cybercriminals to spreadmalicious content. I study the modus operandi of the cybercriminalscontrolling such accounts, and I then develop a system to automatically flag asocial network accounts as fake. I then look at the problem of legitimateaccounts getting compromised by miscreants, and I present COMPA, a system thatlearns the typical habits of social network users and considers messages thatdeviate from the learned behavior as possible compromises. As a last example, I present EvilCohort, a system that detects communities ofonline accounts that are accessed by the same botnet. EvilCohort works byclustering together accounts that are accessed by a common set of IP addresses,and can work on any online service that requires the use of accounts (socialnetworks, web-based emails, blogs, etc.)